Midtraining Research Engineer (AI)

Location: Menlo Park, California, United States; on-site

Salary: $250,000–$350,000 per year plus equity

Company

Periodic Labs is an AI and physical sciences company developing models to accelerate breakthroughs in materials, energy, and scientific discovery.

What you will do

• Identify, process, and curate scientific data for large-scale model training.
• Generate synthetic data to address gaps in scientific knowledge and reasoning.
• Build evaluations that correlate with downstream scientific task performance in collaboration with RL researchers, physicists, and chemists.
• Develop self-distillation and on-policy distillation techniques to improve model capabilities.
• Design and run large-scale training experiments across thousands of GPUs with supercompute engineers.
• Build tools to investigate how data choices affect model intelligence.

Requirements

• Experience training LLMs on curated mixtures containing trillions of tokens.
• Experience supporting large production training runs through a dedicated evaluations team.
• Hands-on experience with self-distillation, on-policy distillation, or similar methods in a real training pipeline.
• Experience with scaling laws and compute-optimal hyperparameters.
• Ability to work across data, evaluations, and training infrastructure.
• Bachelor's degree or equivalent experience.

Nice to have

• Experience optimizing throughput and reliability for large-scale distributed training.
• Background in AI for science or training on specialized datasets such as protein or materials data.
• Experience creating evaluations or synthetic data for non-verifiable tasks and tracking performance during live runs.

Culture & Benefits

• Work alongside RL researchers, physicists, chemists, and supercompute engineers.
• Operate in a rapidly growing environment focused on frontier scientific research.
• Equity is included in the compensation package.
• Visa sponsorship is available, with assistance throughout the process.